MOTM

Rather than have a whole polyphonic DX-7 in a 1U space,
I'd like to see a DX-7 operator in a 3U or 4U space with all the 
control functions handled by individual knobs and switches.  No 
menu's please!  With a number of these you could build a DX-7 voice 
with the patching serving as the algorithm selection.  You could 
patch in filters and other processors between operators within the 
algorithm as well.  I would buy 6 or more without question.  Maybe 
Zimmer would even buy some.  :)

To me, one of the main attractions of a modular is the user 
interface.  I love the one knob equals one function concept.  I 
absolutely hate having to step through multiple layered menus while 
squinting at a tiny LCD screen.  If we can't have real knobs, lets 
have virtual knobs on a large computer screen.  (I'm also one of 
those old guys who can't see quite right anymore.)  

I would also agree that one of the most useful modules would be a 
high quality delay line with a large range.  The range should go 
down perhaps as low as 100 microseconds and perhaps as high as 30 
seconds.  Over at least some of that range you would want 1V/Octave 
control with good precision.  (In other words the delay time would 
half for each octave you went up)  The low range would be useful for 
experimenting with correlation effects and physical modeling, the 
mid range for flanging type effects and the high range for 
simulating tape delay.  The feedback loop should be patchable to 
allow for the insertion of limiting or filtering.  The key to the 
whole thing is to either have no stepping or steps small enough to 
prevent zippering.

Another interesting module could be a filter bank or parametric type 
filter of the FIR type (finite impulse response.)  

How about a fourier series generator?

The possibilities are really endless.  But I personally would rather 
have a large panel with a lot of knobs then a little panel with an 
LCD and layered menus.

Paul Haneberg